1. Field of the Invention
The present invention relates to an error control method and an error control device for digital communication which employ retransmission control such as ARQ (Automatic Repeat Request), and in particular relates to an error control method and an error control device for digital communication which use hybrid ARQ which carries out code combining.
2. Background
In conventional retransmission control using a code combining technique, all received signals are used in a decoding. The technique of retransmission control is disclosed, for example, in IEEE Trans. Commun. vol.33, pp,385-393, May, 1985. Retransmission control methods using code combining techniques are classified into methods which add only an error detecting code to the first signal transmitted without providing a redundancy for error correction, and methods which provide a redundancy for error correction to the first signal transmitted.
In retransmission control using the code combining technique, when signals having an extremely large number of errors are included in the received signals, multiple retransmission is necessary and all the received signals are used in the technique. Furthermore, in the method which adds an error detecting code only to the first transmitted signal without providing a redundancy for error correction, the first signal received can be decoded only if it is received error-free. However, if there is an error in even one bit, then this signal cannot be decoded. On the other hand, in the method which provides a redundancy for error correction to the first transmitted signal, the throughput is improved in the case where the transmission channel has a poor quality. However, in the case where the transmission channel has a sufficient quality, the throughput plateaus.
If Chase decoding is used on retransmission control using a code combining technique, it is possible to correct errors using only an error detecting code. However, many combinations must be calculated in order to carry out Chase decoding, such that the number of calculations becomes enormous. Because in the algorithm defined by Chase decoding, all test patterns are tested, the number of calculations is enormous. As illustrated in FIG. 3, in order to reduce the number of calculations, the Chase decoding algorithm is modified (we define it as modified Chase decoding) such that the trials start with the most reliable test pattern to the most unreliable test pattern and the trials are terminated upon the first successful binary decoding.